According to MotionBeat Monitoring, Kimi Code has officially launched K2.7 Code HighSpeed. This mode has completed the beta phase and is now a permanent option.
The HighSpeed version uses the same model as the regular version. In typical programming scenarios, the output is around 180 Tokens/s, with short contexts reaching up to 260 Tokens/s.
Allegretto and higher tier members do not need to apply and can directly use it in Kimi Code CLI or other tools that support the Coding Plan. CLI users can input /model, select kimi-for-coding-highspeed, and enable it.
The HighSpeed version consumes three times the usage of the regular version, making it more suitable for coding tasks that prioritize response speed.
